Here is a list of my teaching experiences as a PhD student in the department of statistical science at Duke University.

STA790-1: Spatial Topics: Latent Process Modeling

STA199: Introduction to Data Science and Statistical Thinking

STA199: Introduction to Data Science and Statistical Thinking

STA440: Case Studies in the Practice of Statistics

STA432: Theory and Methods of Statistical Learning and Inference

STA250: Statistics

The increasing popularity of statistics or data science in society has attracted many students from a wide range of background to statistics courses. Hence, educators in statistics now face a challenge of accommodating diverse students with different levels of understanding and expectations. When teaching an introductory course to data science (STA199) at Duke University, I had a relatively small but heterogeneous group of students. They spanned multiple majors, student-athletes, different undergraduate years, with a genuine interest in statistics, or with a motivation to simply pass the course. Interestingly however, students commonly have some fears in statistics, which poses another challenge to educators. I have been developing my teaching strategies to address these challenges and foster a motivating and supportive classroom environment where students become active contributors.

First, I think stimulating students’ interest throughout a course is crucial to successfully lead a course until the end. My strategy is to provide “fun” data relevant to students’ everyday life, e.g., Airbnb rentals or airplane delays, or current social issues. For instance, in May 2022, there were major mass shootings, and I made exam questions with the most up-to-date gun violence data. After the exam, students were impressed by how the exam addressed the timely issue and mentioned they “couldn’t get enough of it”. I happily put these efforts so that students discover statistics everywhere and eventually overcome obscure fears that stop them from enjoying statistics. To enhance students’ understanding of statistical concepts, I offer multiple expositions to increase the chance that any one of them resonates with students. It can be a simple formula on the board, visualizations, analogies, anecdotal examples, research or news articles, or some physical illustrations. I also provide multiple layers of learning through different media to bolster students’ understanding. For instance, the R package tidymodels provides a simple one-line code to compute a p-value. Then students are asked to confirm the result by manual calculation from the null distribution of a statistic. In doing so, they obtain clear ideas on what a sample statistic is, what hypotheses they consider, what distribution the statistic follows under the null hypothesis, and eventually what the p-value means. This way, students can go beyond a robotic usage of an R syntax and prepare themselves with a full toolkit applicable to different software and situations.

As a statistics educator, I emphasize the importance of contextualized interpretation of statistical results. In my experience, students are often so concentrated on conducting statistical analyses that they stop after deriving some numbers. Then I bring “so what?” questions so that students learn how to contextualize the result, and communicate it effectively to laypeople. For each newly introduced analysis, students are provided with a plethora of example scenarios where they can practice putting inferential results in context.

Teaching is exciting and challenging because it is so dynamic and creates different chemistry every time. It requires frequent reflection and continuous learning. In particular, I believe adaptivity in teaching styles based on students’ needs is important. From the mid-course evaluation of STA199, I observed obvious separation among students. About half of the students felt the pace was too fast. I had to find the right balance to help everyone successfully complete the course. Hence, I prepared challenging problems that were optional for fast-learning students. These problems kept more advanced students motivated and afforded me some time to go through basic problems with other students more slowly.

To polish my teaching skills, I am enrolled in Certificate in College Teaching program (CCT) at Duke University. This program provides systematic pedagogical training to prepare graduate students as future educators. I have learned how to best serve students in their learning process through 1) offering multiple channels, e.g., emails, office hours, and forums to interact with me, 2) providing ample opportunities to communicate with other students, and 3) establishing clear and transparent expectations via syllabus, rubrics, and written assignment instructions. Graduate students in CCT are required to participate in a teaching observation triad where each student observes and is observed by two peer educators while teaching. Although nerve-wrecking, I find presenting my teaching to other educators highly beneficial because I was able to increase student participation in STA199 based on the peer educators’ advice. I will continue learning from peer educators, my students, and my experiences to become a better educator who is able to accommodate various students, encourage healthy interactions and contributions, and flexibly modify teaching strategies as needed.

I provide a few direct quotes below from student evaluations for STA199.